The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external codebases and workflows to perform refactoring tasks. This processing of untrusted external data presents a surface for indirect prompt injection attacks where malicious instructions embedded in source code could attempt to manipulate the agent.
Ingestion points: The agent is instructed to "Map the workflow boundary" and "identify the concrete pain/failure point" in the user's workspace, which involves reading external files and logs (SKILL.md).
Boundary markers: There are no explicit instructions to use delimiters or ignore instructions embedded within the analyzed code content.
Capability inventory: The skill possesses the capability to modify files within the workspace, as indicated by the workspace-write sandbox mode and the instruction to "Implement or recommend the smallest coherent intervention."
Sanitization: No specific sanitization, escaping, or validation mechanisms are defined for handling the content of the files being refactored.

refactoring-specialist